A Helm chart for Kubernetes
| Key | Type | Default | Description |
|---|---|---|---|
| dataset.local.dataset_path | string | "nil" |
Host Path to input dataset |
| dataset.local.docvqa_path | string | "nil" |
Host Path to docvqa dataset |
| dataset.local.es_data_path | string | "nil" |
Host Path to store es data |
| dataset.local.output_path | string | "nil" |
Host Path to store output |
| dataset.local.postgre_data_path | string | "nil" |
Host Path to store postgres data |
| dataset.local.config_path | string | "nil" |
Host Path to Config Directory |
| dataset.nfs.dataset_path | string | "nil" |
Path to input dataset |
| dataset.nfs.docvqa_path | string | "nil" |
Path to docvqa dataset |
| dataset.nfs.config_path | string | "nil" |
Path to Config in Local NFS |
| dataset.nfs.path | string | "nil" |
Path of NFS |
| dataset.nfs.server | string | "nil" |
Hostname of NFS Server |
| dataset.type | string | "<local/nfs>" |
local,nfs dataset input enabler |
| image.name | string | "intel/ai-workflows" |
|
| metadata.name | string | "document-automation" |
|
| model_name | string | "my_dpr_model" |
|
| proxy | string | "nil" |
|
| pvc.storage | string | "15Gi" |
Size of PVC storage, if the default value cannot meet the requirements, please adjust as needed |
| retrieval_method | string | "ensemble" |
Support ensemble,bm25,dpr |